Accessing watsonx Orchestrate from Bob via CLI

2026-06-15 · Source: Niklas Heidloff · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Robotics & Autonomous Systems · Depth: Intermediate, short

Summary

IBM Bob, a generic AI-based software development tool, can access IBM watsonx Orchestrate environments using an agentic skill and the "orchestrate" CLI. This approach offers several benefits over traditional MCP tools, including better performance for Large Language Models with CLIs due to extensive training data, automatic CLI installation via skills, improved connectivity, and faster updates. The author demonstrated this by reimplementing watsonx Orchestrate ADK MCP Server functionality using a skill and the "orchestrate" CLI. An example flow shows Bob activating a skill, installing the "orchestrate" CLI (version ibm-watsonx-orchestrate==2.10.0), loading dynamic instructions, and executing commands to list deployed agents. The skill is optimized for the local Orchestrate Developer Environment and uses ADK version 2.0.10. CLI documentation is modularized into multiple markdown files to manage context efficiently.

Key takeaway

For AI Engineers integrating IBM Bob with watsonx Orchestrate, consider implementing agentic skills for CLI access instead of MCP tools. This approach ensures automatic CLI installation, better connectivity, and access to the latest features. You should modularize CLI documentation within skills to optimize context loading for agents. This method improves agent reliability and keeps your automation aligned with current "orchestrate" CLI capabilities.

Key insights

Using agentic skills with CLIs offers advantages over MCP tools for AI development agents like IBM Bob.

Principles

• LLMs excel with CLIs due to extensive training data.
• CLIs often update faster than MCP servers.
• Modularize documentation for agent context management.

Method

Implement watsonx Orchestrate ADK MCP Server functionality via an agentic skill. Bob dynamically loads SKILL.md, installs "orchestrate" CLI (version 2.10.0), loads reference docs, and executes CLI commands.

In practice

• Configure Bob skills for automatic CLI setup.
• Break down CLI docs into smaller, referenced files.
• Use "orchestrate agents list" to query deployments.

Topics

• IBM Bob
• watsonx Orchestrate
• CLI Automation
• Agentic Skills
• LLM Integration
• DevOps

Code references

• IBM/ibm-watsonx-orchestrate-adk
• nheidloff/orchestrate-bob

Best for: AI Engineer, Machine Learning Engineer, MLOps Engineer

Related on AIssential

• Watsonx Orchestrate Debug Skill for IBM Bob — IBM Bob's custom debug skill automates local watsonx Orchestrate agent issue identification and resolution.
• One Flexible Tool Beats a Hundred Dedicated Ones — Flexible CLI tools outperform dedicated MCP tools for LLM agents due to improved compositionality and reduced context bloat.
• Developing watsonx Orchestrate Agents with Antigravity — Google Antigravity can build IBM watsonx Orchestrate agents using open standards like MCP.
• The Worst Coder in the World goes agentic: building a leaderboard cracking AI — AI agents, powered by standardized protocols like MCP, significantly lower the barrier to entry for software development and knowledge sharing.
• CLI vs MCP: How AI Agents Choose the Right Tool for the Job — CLI excels for known commands, while MCP provides necessary abstraction and governance for complex tasks.
• How to design and run an agent in rehearsal – before building it — DataRobot Agent Assist accelerates AI agent development from design to deployment via a natural language CLI.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Niklas Heidloff.